DEPARTMENT OF CHEMISTRY
ATOMIC AND MASS SPECTROMETRY (A&MS) RESEARCH UNIT
METHOD DEVELOPMENT FOR SPENT NUCLEAR FUEL
CHARACTERISATION USING ISOTOPE DILUTION
HPIC-SF-ICP-MS
Nancy Wanna
Student number: 01613721
Supervisor & promotor: Prof. Dr. Frank Vanhaecke Mentor: Dr. Karen Van Hoecke
Co-promotor: Dr. Andrew Dobney Co-mentor: Dr. Mirela Vasile
A dissertation submitted to Ghent University in partial fulfilment of the requirements for the degree of Doctor of Science:
Chemistry
Academic year 2020–2021
For my family
Members of examination committee
I. Voting members:
Chair:
Prof. Dr. Mieke Adriaens Ghent University
Department of Chemistry - WE06 Secretary:
Dr. Stepan Chernonozhkin Ghent University
Department of Chemistry - WE06 Other voting members:
Dr. Laura Aldave De Las Heras
European Commission’s Joint Research Center JRC Karlsruhe Institute of Nuclear Safety and Security
Dr. Hélène Isnard
French Atomic Energy Commission CEA Saclay Department of Physico-Chemistry
Prof. Dr. Anna Kaczmarek Ghent University
Department of Chemistry - WE06 Prof. Dr. Frederic Lynen
Ghent University
Department of Organic and Macromolecular Chemistry - WE07 Prof. Dr. Laszlo Vincze
Ghent University
Department of Chemistry - WE06 II. Non-voting members:
Prof. Dr. Frank Vanhaecke Ghent University
Department of Chemistry - WE06 Dr. Andrew Dobney
Belgian nuclear research center SCK CEN Institute of Nuclear Materials Science
Table of contents
Acknowledgements ... i
Preface ... iii
Abbreviations ... v
Units used ... viii
Chapter 1 – Introduction and objectives ... 1
1.1 Spent nuclear fuel ... 1
1.2 Radioactive waste management ... 3
1.3 Spent nuclear fuel management ... 4
1.4 Spent nuclear fuel analysis ... 6
1.4.1 Uranium ... 8
1.4.2 Plutonium ... 8
1.4.3 Lanthanides ... 9
1.4.3.1 Neodymium ... 9
1.4.3.2 Gadolinium ... 10
1.4.4 Other spent fuel components ... 11
1.4.5 Non-Destructive analysis ... 12
1.4.6 Destructive analysis ... 13
1.4.6.1. Radiometric methods ... 15
1.4.6.2. Mass spectrometry methods ... 16
1.5 Aim and objectives ... 19
1.6 References ... 20
Chapter 2 – Ion chromatography ... 24
2.1 Introduction and overview ... 24
2.2 Ion exchange chromatography theory ... 24
2.2.1 Common terms in chromatography ... 25
2.3 Components of HPIC systems ... 27
2.3.1 Eluents ... 27
2.3.2 Pump system ... 28
2.3.3 Sample injection unit... 29
2.3.4 Stationary phases and columns ... 30
2.3.5 Detectors ... 31
2.3.5.1 UV-Vis detectors ... 31
2.4 Nuclear applications with HPIC ... 34
2.4.1 Nuclearized HPIC-SF-ICP-MS setup ... 35
2.5 Conclusion ... 37
2.6 References ... 40
Chapter 3 – Inductively coupled plasma-mass spectrometry ... 44
3.1 Introduction and overview ... 44
3.2 Background on SF-ICP-MS ... 45
3.2.1 Operating principle ... 45
3.2.2 Isotope ratio measurement ... 45
3.2.3 Instrumental mass bias ... 46
3.2.4 Dead time ... 48
3.2.5 Spectral interferences ... 49
3.2.6 Quantification methods ... 49
3.2.6.1 External calibration ... 50
3.2.6.2 Standard addition ... 50
3.2.6.3 Isotope dilution ... 51
3.2.6.4 Internal normalization ... 54
3.3 Components of single-collector SF-ICP-MS ... 54
3.3.1 Sample introduction system ... 55
3.3.2 Inductively coupled plasma ... 57
3.3.3 Interface ... 59
3.3.4 Ion transfer optics ... 59
3.3.5 Slit system ... 60
3.3.6 Sector field mass separator ... 61
3.3.7 Detection system ... 63
3.3.8 Vacuum system ... 64
3.4 Nuclear application of ICP-MS... 65
3.4.1 Nuclearization of a single-detector double-focusing SF-ICP-MS ... 66
3.5 References ... 70
Chapter 4 – HPIC-SF-ICP-MS separation method development and validation 75
4.1 Introduction ... 75
4.2 Methods ... 76
4.2.1 Setup timings ... 76
4.2.2 SF-ICP-MS acquisition parameters ... 77
4.2.3 HPIC-SF-ICP-MS separation method development ... 78
4.2.3.1 Separation of lanthanides ... 78
4.2.3.2 Separation of uranium and lanthanides ... 78
4.2.3.3 Separation of plutonium, uranium and lanthanides ... 78
4.2.4 HPIC-SF-ICP-MS separation method validation ... 79
4.2.4.1 Linearity ... 79
4.2.4.2 Repeatability and intermediate precision ... 79
4.2.4.3 Limit of detection (LOD) & Limit of quantification (LOQ) ... 79
4.2.5 Other matrices ... 80
4.2.5.1 Spent nuclear fuel matrix ... 80
4.2.5.2 Environmental soil matrix ... 80
4.3 Results and discussion ... 80
4.3.1 Setup timings ... 80
4.3.2 HPIC-SF-ICP-MS separation method development ... 81
4.3.2.1 Separation of lanthanides ... 81
4.3.2.2 Separation of uranium and lanthanides ... 84
4.3.2.3 Separation of plutonium, uranium and lanthanides ... 85
4.3.2.3.i Oxidation of plutonium ... 85
4.3.2.3.ii Elution of plutonium, uranium and lanthanides ... 86
4.3.3 HPIC-SF-ICP-MS separation method validation ... 90
4.3.3.1. Linearity ... 90
4.3.3.2. Repeatability and intermediate precision ... 90
4.3.3.3. LOD and LOQ ... 91
4.3.4 Other matrices ... 92
4.3.4.1 Spent nuclear fuel matrix ... 92
4.3.4.2 Environmental soil matrix ... 93
4.4 Conclusions and outlook ... 94
4.5 References ... 96
Appendix chapter 4 ... 99
Chapter 5 – Optimization of isotope dilution HPIC-SF-ICP-MS ... 112
5.1 Introduction ... 112
5.2 Materials and methods ... 112
5.2.1 Standards and spikes ... 112
5.2.2 Fuel samples ... 113
5.2.3 Optimization of IR precision in transient signals... 113
5.2.3.1 HPIC method for Nd ... 113
5.2.3.2 SF-ICP-MS method for Nd... 113
5.2.3.3 Calculation methods of IRs from transient signals ... 114
5.2.3.3.i Point by point method ... 115
5.2.3.3.ii Linear regression slope method ... 115
5.2.3.3.iii Peak area integration method ... 116
5.2.3.4 Calculation methods of IRs from continuous nebulization ... 116
5.2.4 Isotope dilution ... 117
5.2.4.1 Instrumental methods ... 118
5.2.4.1.i HPIC method... 118
5.2.4.1.ii SF-ICP-MS method ... 118
5.2.4.1.iii TIMS and alpha-spectrometry method ... 119
5.2.4.2 Calculation methods ... 120
5.3 Results and discussion ... 123
5.3.1 Optimization of IR precision in transient signals... 123
5.3.2 Isotope dilution ... 131
5.3.2.1 UOx fuel ... 131
5.3.2.2 Gd fuel ... 135
5.4 Conclusions and outlook ... 137
5.5 References ... 138
Appendix chapter 5 ... 140
Chapter 6 – IDMS data analysis using R ... 143
6.1 Introduction ... 143
6.2 Methods ... 143
6.2.1 Plotting the chromatographic peaks ... 143
6.2.3 Determination of isotopic composition and mass fraction of Nd using R ... 144
6.3 Results ... 144
6.3.1 Plotting the chromatographic peaks ... 144
6.3.2 Calculating IRs ... 145
6.3.3 Determination of isotopic composition and mass fraction of Nd using R ... 145
6.4 Conclusions and outlook ... 146
6.5 References ... 147
Chapter 7 – General conclusions and outlook ... 148
7.1 General conclusions ... 148
7.2 Further investigations ... 149
7.2.1 Spent nuclear fuels ... 149
7.2.2 Environmental samples ... 149
7.2.2.1 Dealing with matrix composition of environmental samples ... 149
7.2.2.2 Improving the LOD for Pu in environmental samples ... 150
7.2.2.3 Fast determination of U and lanthanide concentrations using external calibration or single standard addition ... 150
Summary ... 151
Samenvatting ... 155
List of publications ... 159
Acknowledgements
This PhD unraveled a world of opportunities for me to learn, grow and challenge myself in various ways that I can look back at with contentment. I am grateful to have been given the opportunity to perform research in the nuclear field and to share my work with the world.
Therefore, I would like to thank the people who made all of this possible. First, I would like to thank my university supervisor and promotor Prof. Dr. Frank Vanhaecke for the opportunity to perform this PhD research and for his guidance and support. Since the day we met at Ghent University, I have been looking up to you as a guru of ICP-MS and at the same time as a friendly and down-to-earth person.
I would like to thank my mentor Dr. Karen Van Hoecke for her devotion, guidance, support and kindness. I will not forget the enormous amount of time and effort that you dedicated to make sure that everything is “going on the right track”, starting from the 13th of January, 2017 when you met me at Brussel’s airport with a train schedule and map, until the day this PhD is finalized. You are a role model of professionalism and leadership and the memories we shared together at work and in our respective personal lives will always be special to me.
I would like to thank my co-promotor Dr. Andrew Dobney for his support and for sharing his expertise with me throughout this PhD. It has been a pleasure learning from you not only about work, but about different aspects of life as well. Thank you for making time for me, my PhD and my life as an expat living in Belgium, for sharing the ups and downs with me throughout these four years and for motivating me whenever I was faced with difficulties along the way.
I am grateful for the support and guidance of my co-mentor Dr. Mirela Vasile. You have contributed to this PhD not only with time, effort and expertise about environmental samples but also with friendliness and kindness which made the work more enjoyable. Your time and people management skills are inspiring, and I look up to the way you give out everything in your power to reach your goal.
I would like to thank the Academy at the Belgian nuclear research center (SCK CEN) and especially Dr. Michèle Coeck and the scientific jury members for this opportunity of a life time.
Also from SCK CEN, I would like to thank Prof. Dr. Thomas Cardinaels, Dr. Ingrid Geuens and Dr. Michel Bruggeman for including me in their respective groups: radiochemistry group (RCY), radiochemical analyses group (RCA) and low-level radioactivity measurements group (LRM). I am grateful to colleagues and supporting staff members of the RCA group, especially Dr. Lesley Adriaensen, Peter Van Bree, Els Verheyen, Luc Gelens, Prisca Verheyen, Karolien Van Rompaey, Loes Van Hout, Göran Verpoucke and Simon Sauvillers for the enjoyable work atmosphere, sharing their expertise, ensuring the safety of the work performed with radioactive samples and for their help and support throughout the four years of this PhD. I would also like to thank staff members of the RCY group especially Dr. Peter Zsabka, Dr. Karen Van Hecke and Dr. Mireille Gysemans for the enjoyable work atmosphere and friendly conversations, and staff members of the LRM group especially Karin Jacobs, Diana Verstrepen and Anke Hooyberghs for their time and help in preparing the environmental samples as well as for their
friendliness and kindness. I could not have written the R code without the help and support of Dr. Bart Rogiers, thank you for teaching me everything I know about R.
Since my first days at SCK CEN, I have met Yaana Bruneel and Dr. Yana Dekempeneer, who are both inspiring, supportive, kind and smart women that I am lucky to have as friends. I would like to thank fellow PhD students at the chemistry building of SCK CEN for the friendly conversations, the fruitful discussions and for sharing their expertise whenever needed.
I am grateful to helpful staff members of the atomic and mass spectrometry group at Ghent University, especially Dr. Stepan Chenonozhkin and Kris Latruwe for sharing their expertise on MC-ICP-MS and for the measurement of standards.
I would like to acknowledge the beautiful, supportive, life-loving and kind-hearted friends and family members (in Lebanon and all over the world) who were far away in distance but have never failed to support and encourage me no matter where they are. Thank you for keeping me in your prayers, for bringing me joy no matter the circumstances and for being my piece of home away from home.
A special gratitude goes to my dear husband Alexander. During these last four years, nothing would have been the same without you in my life. Thank you for your companionship and for enduring this PhD journey with me. I am looking forward to discovering together what the future holds for us. Life has been full of surprises and I have always been inspired by my two
“little” brothers, Khalil and George, who can handle any situation without fear or panic. Thank you both, for your sense of humor, kindness and support. And last but undoubtedly not least, I would like to express my lifelong gratitude to my parents, Nazem and Antoinette, who have loved me unconditionally and supported all my decisions and journeys in life, even when it meant that I had to be thousands of kilometers away from them. The happiness and wellbeing of their children, is always their priority and it came at the expense of their health, energy and comfort. They have given me and my brothers everything we could ever need and more, and not a day goes by without me being grateful to God for giving me such amazing parents. I dedicate this dissertation to my parents and I hope I can always make them proud, to make their countless efforts worthwhile.
Preface
This PhD project was performed at facilities on the site of the Belgian Nuclear Research Centre SCK CEN (Mol, Belgium) in collaboration with Ghent University (Ghent, Belgium), and was focused on developing an analytical method to characterize spent nuclear fuel (SNF) using high pressure ion chromatography (HPIC) coupled to a double focusing single collector sector field inductively coupled plasma-mass spectrometer (SF-ICP-MS).
Determining the isotopic and elemental composition of SNF, that is fuel which has undergone irradiation in a nuclear reactor, is essential for nuclear waste management and for evaluating the performance of the fuel. Post-irradiation examination of SNF includes destructive analysis for the determination of the mass fractions of long-lived actinides and lanthanides (fission products). The long-lived or stable nature of the nuclides of interest makes ICP-MS a more suitable method for their analysis and the method of choice for this work, rather than counting for example by using alpha-spectrometry. Isotope dilution mass spectrometry can be used as a primary method for determining mass fractions of uranium (U), plutonium (Pu) and neodymium (Nd), with the best precision and accuracy based on isotope ratio (IR) measurements using ICP- MS or, alternatively, by thermal ionization mass spectrometry (TIMS) in combination with alpha-spectrometry (e.g. for 238Pu). SF-ICP-MS is a single detector ICP-MS instrument well known for its flat top spectral peaks at low mass resolution, which increase the precision of IR measurements compared to quadrupole ICP-MS. Coupling HPIC on-line with ICP-MS (not possible with TIMS and alpha-spectrometry) offers a fast, and precise way to eliminate notorious isobaric interferences whilst reducing the analyst’s exposure to radiation. The literature reports HPIC coupling to various types of ICP-MS instruments for the characterization of SNF. However, the overall precision obtained with isotope dilution when using HPIC coupled with a single detector SF-ICP-MS to determine the concentrations of U, Pu, Nd and Gd in SNF from one injection has not been reported previously. Therefore, the aim of this work was to separate U, Pu, Nd and Gd and to determine their mass fractions in SNF by means of isotope dilution HPIC-SF-ICP-MS. To accomplish this, the work was divided into four objectives as follows:
(1) Develop and validate a separation method for U, Pu and the lanthanides from one injection using HPIC-SF-ICP-MS. The first objective was tackled in different steps where, at each step, simulations with the Hydra/Medusa software package were performed to predict the complexes in solution. All Pu species in the sample had to be oxidised into Pu(VI) prior to the injection of the sample onto the column.
(2) Optimize the acquisition parameters with the SF-ICP-MS set-up and select a calculation method to obtain the most precise IR measurements. This optimization was accomplished by investigating different acquisition parameters with repeated injections of a Nd standard of natural isotopic abundance. Four mass windows (2, 25, 50 and 150 %), two dwell times (10 and 30 ms) and two different numbers of isotopes monitored per run (2 and 7) were investigated.
At the same time, the three IR calculation methods most commonly reported in the literature were compared for their accuracy and precision, namely linear regression slope (LRS), point by point (PbP) rationing and peak area integration (PAI).
(3) Characterize two types of SNF (UOx and “Gd fuel”) using isotope dilution HPIC-SF-ICP- MS. Isotope dilution parameters (other than those investigated in the second objective) had to be optimized for on-line determination of Pu, Nd and Gd elemental mass fractions and a separate off-line determination of the U mass fraction in two types of SNF. This optimization included determining the lowest error magnification factor to obtain the most precise IR in the blend.
(4) Determine the overall uncertainty budget of isotope dilution HPIC-SF-ICP-MS, to compare this developed method to the ISO 17025 accredited isotope dilution TIMS & alpha- spectrometry method, which is currently used at SCK CEN. Uncertainty calculations were performed using the GUM Workbench software.
Abbreviations
PAR 4-(2-pyridylazo) resorcinol α-HIBA Alpha-hydroxyisobutyric acid
ASTM American Society for Testing and Materials SCK CEN Belgian Nuclear Research Centre
BP Binary pump
CE Capillary electrophoresis
CEA Commissariat à l’Energie Atomique et aux énergies alternatives CCQM Consultative Committee for Amount of Substance
CITAC Co-operation on International Traceability in Analytical Chemistry DAD Diode array detectors
DCP Direct current plasma EDF Electricité de France ESA Electrostatic analyser
EGADSNF Expert Group on Assay Data for Spent Nuclear Fuel EGBUC Expert Group on Burn-Up credit Criticality
FP Fission product
GC Gas chromatography
GDMS Glow discharge mass spectrometry HETP Height equivalent to a theoretical plate HPIC High pressure ion chromatography HPLC High pressure liquid chromatography HPGe High purity germanium
HKED Hybrid K-edge densitometry
ICP-MS Inductively coupled plasma-mass spectrometry ISO International Organisation for Standardisation IAEA International Atomic Energy Agency
BIPM International Bureau of Weights and Measures OIML International Organization of Legal Metrology
IC Ion chromatography
IEC Ion exchange chromatography
IUPAC International Union of Pure and Applied Chemistry IDMS Isotope Dilution Mass Spectrometry
IR Isotope ratio
JRC Joint Research Centre LOD Limit of detection LOQ Limit of quantification LRS Linear regression slope LSC Liquid scintillation counting K-bias Mass bias correction factor
R Mass resolution
MOx Mixed oxide
MC Multi-collector
NIST National Institute of Standards and Technology
NIRAS Nationale Instelling voor Radioactief Afval en Verrijkte Aplijtstoffen NRD Neutron resonance densitometry
NEA Nuclear Energy Agency
ORNL Oak Ridge National Laboratory
ONDRAF Organisme National des Déchets Radioactifs et des Matières Fissiles Enrichies PDET Partial defect tester
PIPS Passivated implanted planar silicon PSI Paul Scherrer Institute
PAI Peak area integration
%FIMA Percentage of fissions per initial metal atoms PFA Perfluoroalkoxy
PAD Photodiode array detectors PbP Point-by-point
PEEK Polyether ether ketone PTFE Polytetrafluoroethylene
PVF Polyvinylfluoride
PWR Pressurized water reactors
RF Radiofrequency
REE Rare earth elements RSD Relative standard deviation RSC Royal Society of Chemistry
STUK Säteilyturvakeskus (radiation and nuclear safety authority in Finland) SEM Secondary electron multipliers
SIMS Secondary ion mass spectrometry SF Sector field
SINRD Self-indication neutron resonance densitometry
SP Single pump
SNF Spent nuclear fuel SD Standard deviation
SSB Standard sample bracketing
TIMS Thermal ionization mass spectrometry UV-Vis Ultraviolet-Visible
UOx Uranium oxide
VWD Variable wavelength detectors
Units used
Name Symbol Description
Bar bar Unit of pressure equal to 100,000 Pa.
Barn b Unit used in particle physics to express the cross sectional area of nuclei. A barn is equal to 10-28 m2.
Becquerel Bq SI unit of the activity referred to a radionuclide. Becquerel is the rate of radioactive decay of an unstable nuclide and is equal to one reciprocal second.
Electron volt eV Kinetic energy acquired by an electron in passing through a potential difference of one volt in vacuum. An eV is equal to 1.602176634 · 10-19 J.
Equivalent equiv Unit for an amount of a substance in solution and is equal to the number of moles of an ion in solution multiplied by the valence of that ion.
Kelvin K SI unit of thermodynamic temperature and is equal to 1/273.16 of the triple point of water.
Megawatt electric
MWe One million watts of electric capacity
Pascal Pa SI unit of pressure and is equal to newton per square meter.
Sievert Sv SI unit for dose equivalent and is equal to the joule per kilogram.
Unified atomic mass unit
u Unit equal to 1/12 of the mass of a free carbon 12 atom, at rest and in its ground state.
Chapter 1 – Introduction and objectives
This first chapter provides a general overview of spent nuclear fuel (SNF), its management and characterization methods. The aim and objectives of this work will also be outlined in this chapter.
1.1 Spent nuclear fuel
SNF is part of the radioactive waste generated in nuclear power plants. According to the International Atomic Energy Agency (IAEA), SNF refers to fuel assemblies discharged from nuclear reactors after irradiation [1]. The nuclear fuel currently used in most reactors is based on uranium oxide with different enrichment values, depending on the reactor type. For example, the fuel used in light water reactors, such as pressurized water reactors (PWR) (Figure 1.1) and boiling water reactors, is enriched in 235U up to 5 wt. %, while pressurized heavy water reactors use natural (~ 0.7 wt. % 235U) or slightly enriched (up to 1.2 wt. % 235U) uranium. After its mining, uranium is converted to uranium hexafluoride, which is in gaseous form, to be enriched [2]. After enrichment, uranium hexafluoride is converted to uranium dioxide (by “dry” or “wet”
chemical processes [2]), which is subsequently compressed into fuel pellets (10 mm in diameter and 10-15 mm in height) which are then stacked into long tubes around 3-5 m in length, made of Zircaloy to make fuel rods for PWRs [3]. Zircaloy is an alloy of more than 95 wt. % zirconium, which has a very low absorption cross-section for thermal neutrons, high hardness and is durable and corrosion-resistant. Fuel rods (as many as 150 to 260) are grouped together into a specific geometry (square or hexagonal cross-section) using spacers to form a fuel assembly (Figure 1.1). Typically, a 1000 MWe PWR contains between 120 and 200 fuel assemblies. As shown in Figure 1.1, in a PWR, the reactor and steam generators are contained in a reinforced concrete structure (concrete and stainless steel) to protect these components from outside elements and to contain the radiation in case of any major accident.
Figure 1.1 Schematic representation of a PWR (left) [4] and a PWR fuel assembly (right) [2]
When struck by a thermalized neutron, 235U becomes unstable and undergoes fission. During fission, the 235U atom splits into two smaller atoms, called fission products, having masses of around 95 and 140 atomic mass units (Figure 1.2), and releasing heat and 2 to 3 neutrons, which can support a fission chain reaction. To control a fission chain reaction, fissile material, neutrons, and material to slow down/retain neutrons are required. To sustain a fission chain reaction, neutrons are slowed in the reactor core by a moderator, which is usually water, but can be heavy water or graphite. To control or halt a fission chain reaction, neutron absorbers can be used in liquid form, such as boric acid dissolved in the primary water circuit, or in solid form, such as control rods (containing hafnium, boron or cadmium) or a burnable poison with high neutron absorption cross-section (such as gadolinium) can be included in the fuel to limit the excess reactivity. Liquid and solid neutron absorbers can also be used in combination to control the number of neutrons in a nuclear reactor. The neutron multiplication in a fission is characterized by the parameter k, defined in eq. 1.1 where η is the number of neutrons produced in a fission to the number of neutrons absorbed, 𝜖 is the fast fission factor, which is the ratio of the total number of neutrons produced by all fissions to that of slow neutron fission, p is the resonance escape probability, which is the probability that neutrons avoid being captured and reach thermal energies where they may cause fission, f is the thermal utilization, which is the ratio of thermal neutrons absorbed in the fuel to the total number of thermal neutrons absorbed and pfnl and ptnl are the probabilities of fast and thermal neutron non-leakage, respectively [5].
A reactor is critical and has a constant power when k = 1, is supercritical and its power increases when k > 1, and is subcritical and its power decreases when k < 1.
𝑘 =𝑁𝑒𝑢𝑡𝑟𝑜𝑛 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛 𝑟𝑎𝑡𝑒
𝑁𝑒𝑢𝑡𝑟𝑜𝑛 𝑙𝑜𝑠𝑠 𝑟𝑎𝑡𝑒 = 𝜂. 𝜖. 𝑝. 𝑓. 𝑝𝑓𝑛𝑙. 𝑝𝑡𝑛𝑙 (eq. 1.1) Nuclear fuel is considered “spent” when it no longer contains enough 235U atoms to sustain the fission chain reaction. This can happen 3 to 7 years after fuel loading, depending on the fuel and its location in the reactor core. Neutron capture by 238U leads to the formation of plutonium and transuranic elements called minor actinides, such as neptunium, americium and curium, which can also undergo fission. SNF is highly radioactive as it emits alpha, beta, gamma and neutron radiation. The radiation level of SNF decreases over time as the radioactive elements decay. There are three main sources of radiation in SNF. These are (1) radioactive fission products (e.g. in increasing order of half-life from 5.2 days to 2.1·105 years: 133Xe, 131I, 85Kr,
152Eu, 90Sr, 137Cs and 99Tc), (2) products of neutron capture by uranium such as plutonium (e.g.
in increasing order of half-life from 14.3 to 3.7·105 years: 241Pu, 240Pu, 239Pu and 242Pu) and minor actinides (e.g. in increasing order of half-life from 2.4 days to 2.1·106 years: 239Np,
242Cm, 244Cm, 241Am and 237Np), and lastly (3) activation products formed by neutron capture in fuel cladding and structural materials in the fuel assembly (e.g. in increasing order of half- life from 2.7 to 3·105 years: 55Fe, 60Co, 14C, 94Nb, 59Ni and 36Cl). Some components of SNF can be reusable, either for producing fresh fuel (uranium and plutonium) or for other uses, such as irradiation in a medical context (fission product: caesium).
Figure 1.2 Distribution of fission products produced in a UOx fuel [6]
1.2 Radioactive waste management
Radioactive waste can be generated from a wide range of applications in addition to nuclear energy generation, such as the use of radioactive sources in medicine, research, agriculture and industry. The physical, chemical and radiological characteristics of waste vary widely depending on the application, but a common characteristic of all radioactive waste is its potential of being a hazard to people and to the environment. The analysis of radioactive waste is important for radioactive waste management prior to its disposal. National safety regulations are set for managing radioactive waste. However, the risk of radiation may transcend national borders. Therefore, the IAEA issued several international safety standards [7-9] and guides [10- 13] providing recommendations for safe management and storage of radioactive waste to protect people and the environment. Radioactive waste management consists of the following steps: (1) collecting the radioactive waste, (2) processing it into a form suitable for safe storage and (3) storing it in surface or geological repositories depending on its classification. At various steps of the radioactive waste management process, the characterization of the different properties (e.g. physical, chemical and radiological) of radioactive waste is required and recorded to facilitate its management.
Radioactive waste is classified by the IAEA into six categories based on the radioactivity level and half-life (the time it takes for the radioactivity to decrease by half) of the waste [14] and ways for its safe management are indicated by the European Commission in the radioactive waste and spent fuel management directive 2011/70/Euratom [15]. Nevertheless, classification of radioactive waste can vary between countries. In Belgium, for example, NIRAS/ONDRAF (Nationale Instelling voor Radioactief Afval en verrijkte Splijtstoffen / Organisme National des Déchets Radioactifs et des matières Fissiles enrichies), the national organisation in charge of managing radioactive waste, classifies radioactive waste for final disposal into three categories:
A (short-lived low and intermediate level waste), B (long-lived low and intermediate level waste) and C (short and long-lived high-level waste), which are based on the radioactivity content and the half-life of the waste similar to the IAEA classification. In contrast, the radiation and nuclear safety authority STUK (Säteilyturvakeskus) in Finland distinguishes between nuclear waste (spent nuclear fuel & low and intermediate nuclear waste) and radioactive waste
(conditioned solid waste, liquid waste, solid waste and airborne discharges), based on their disposal routes.
European countries have chosen to store their short-lived low and intermediate activity waste differently, by either opting for geological (e.g. Forsmark, Sweden) or near-surface storage (e.g.
Aube, France). However, for long-lived high-level waste there is a general consensus that geological disposal is the safest option. Currently in Belgium, short-lived low and intermediate level waste undergoes intermediate near-surface storage, whilst the final disposal site in Dessel is still under construction according to NIRAS/ONDRAF [16].
1.3 Spent nuclear fuel management
Different strategies for managing SNF exist, since it can be regarded as a resource, and reprocessed to make new nuclear fuel (generating a closed fuel cycle), or it can be considered as radioactive waste after its removal from the reactor core and needing to be disposed of after decades of interim storage (open fuel cycle). The choice of spent nuclear fuel management strategy varies between countries, for example, France, the Russian Federation, China, India and Japan have chosen to reprocess their SNF, while the United States of America, Canada, Finland, Sweden and Germany opted for directly disposing their SNF. If classified as a resource, SNF assembly is dissolved in acid before separating uranium and plutonium from fission and activated products and transuranic elements (vitrified high level waste) and from the cladding and structural materials (intermediate level waste). Dissolving the SNF in acid does not achieve any separation of U and Pu; dissolution merely converts the solid SNF into solution form. The separation takes place after the dissolution, by a variety of processes, one of which is the PUREX process, a solvent extraction based process. SNF dissolves readily in a wide concentration range of nitric acid, with the consumption of protons and nitrate to form uranyl nitrate species. Many different chemical reactions with various stoichiometries can take place [17]. Uranium starts to precipitate from +/- pH 5 onwards under oxidizing conditions, hence only acidic conditions are used. Uranium and plutonium are then separated from each other and the uranium is isotopically enriched. Then, uranium and plutonium are converted into oxides that are used to make fresh mixed oxide fuel (MOx). Compared to the open fuel cycle, reprocessing involves additional steps, but reduces demand for uranium by 25 %. Both SNF and high-level waste require disposal in geological repositories, while intermediate level waste can be stored in near-surface or underground facilities. However, if classified as waste, SNF can be stored in water pools for up to 3 to 4 decades before being transported for interim storage or for disposal in a geological repository. Pools serve two functions: (1) protecting the workers from the radiation emitted by the SNF assemblies and (2) cooling the SNF assemblies by evacuating the heat through the circulating water. In geological disposal, SNF and high-level waste are encapsulated in a leak-tight container, isolated from the environment by multiple barriers and placed in tunnels at several hundred meters of depth. At different stages of SNF management, certain nuclides become more important (Figure 1.3). For example, during transportation and storage of SNF, short-lived radionuclides (such as volatile 131I) are significant in case of damage breaching the fuel rod cladding within one year of its discharge from the reactor core.
Figure 1.3 Trend of the activity of different SNF components relative to mined uranium ore with respect to time [3]
Shielding is a requirement throughout transportation and storage of SNF due to the hazard from
60Co and 137Cs, which are gamma emitters, and 240Pu and 242Cm, which are neutron emitters.
Shielding is provided by 3 to 4 meters of water in pools where SNF is first stored (wet storage) after being discharged from the reactor core, before being transferred to dry storage where shielding is provided by the metal/concrete cask body and neutron absorbing materials. SNF is not moved to dry storage until after at least 5 to 10 years of storage in a reactor pool (depending on its capacity) connected directly to the reactor. Additional away-from-reactor storage might be needed, either wet or dry storage, if the capacity of the reactor pool is reached. Long-term dry storage for up to 100 years is now being considered due to the lack of final disposal facilities. The licence to construct the world’s first geological repository (Onkalo) at the Olkiluoto site was granted in Finland in November 2015 [18] and the application for the operational license is scheduled to be submitted by the end of 2022 [19]. Additionally, cooling of SNF is required due to heat released by the beta emitting radionuclides 137Cs and 90Sr.
Cooling ensures that no temperature limits will be exceeded in operational or accident conditions in order to protect the storage facility’s structures, systems, components and the fuel from damage throughout the lifetime of the storage facility [1], thereby preventing release of radioactive material to the environment. In wet storage, cooling is provided by the pool water that is constantly circulated through heat exchangers, whereas in dry storage, cooling is provided by forced or natural air circulating around the containers of SNF [1]. Transport of SNF might be needed depending on the location of the nuclear power plant and SNF management facility. For transport of SNF, cylindrical containers are used, after first having passed a series of tests (drop, heat, immersion in water, etc.) to prove their robustness and thermoconductive and shielding properties, as recommended by the IAEA [1]. Design of the geological repositories must take into account the long-term heat release of 137Cs, 90Sr, 241Pu and 241Am and the mobility of long-lived radionuclides, such as 99Tc, 14C, 36Cl and 239Np, to ensure safe disposal of SNF. SNF still contains fissile uranium and plutonium radionuclides, which could become critical causing a chain reaction. Subcriticality must be ensured during
transport, storage, reprocessing and in geological repositories, for example, when stored in reactor pools, a subcritical geometry must be maintained and neutron absorbers can be added.
1.4 Spent nuclear fuel analysis
The composition of SNF depends on the amount of energy liberated by the fuel. An important characteristic of SNF is the “burn-up”, which is the number of fissions undergone by the fuel [20] and the clearest way to express it, is as the percentage of fissile metal atoms that underwent fission (%FIMA). This is calculated as shown in eq. 1.2 [21], where M & M’ are the numbers of heavy metal atoms before and after irradiation, F is the number of heavy metals that underwent fission, and is determined most frequently as F = N/Yeff, N being the number of fission product atoms of a burn-up monitor nuclide (such as 148Nd) and Yeff being the effective fission product yield weighted by the yield for each fissile actinide.
%𝐹𝐼𝑀𝐴 = 𝐹
𝑀= 𝐹
𝐹+𝑀′ (eq. 1.2)
The %FIMA is most often calculated using 148Nd, although also other fission products can be used as burn-up monitors such as 137Cs, 139La and 144Ce. In order to use 137Cs for burn-up determination, it is essential to analyse a fuel sample that is large enough to be representative, because Cs is known to migrate from the centre of the fuel pellet towards its edge during irradiation due to the high radial temperature gradient between the pellet core (1200 °C) and rim (400-500 °C) [22]. In practice, the 144Ce half-life of 285 days can limit the use of this nuclide as burn-up monitor to only those SNFs with short cooling times [21]. The following properties of 148Nd make it the most used fission product monitor: (1) it is not volatile, does not migrate in the fuel pellet and has no volatile precursor, (2) it is stable and does not require decay corrections, (3) its fission yield is nearly the same for 235U (1.66 %) and 239Pu (1.65 %) [23], (4) it is not present in fresh nuclear fuel, (5) it can be corrected for natural neodymium contamination using 142Nd (which is stable and not a normal constituent of fresh fuel [24]), (6) it has a low neutron absorption cross-section (Table 1.4), which means that its probability of taking up another neutron is low.
The %FIMA (shown in eq.1.2) can be converted to gigawatt days per metric ton by consulting the ASTM E321 – 96 standard test method for burn-up determination [25]. SNF from UOx fuels with 3.5 % initial enrichment, 33 GWd·t(heavy metals)-1 energy output and 3 years cooling time, contains (on a metals basis) almost 96 wt. % of uranium (including 1 % of fissile
235U), 1 wt. % of plutonium, 0.1 wt. % of minor actinides and 3 wt. % of fission products (Figure 1.4) [1]. For fuel with a higher burn-up, the amount of uranium remaining will be lower than 96 wt. % and the amounts of plutonium, minor actinides and fission products will be higher than 1, 0.1 and 3 wt. %, respectively (on a metals basis). In “Gd fuels”, the fresh fuel is enriched to 5-10 wt. % in gadolinia (Gd2O3). As such, Gd is present as a major element within such fuels (before and after irradiation). Apart from the presence of the additional Gd, the composition of these “Gd fuels” is otherwise comparable to that of UOx fuels. The determination of the isotopic composition and content of Gd in “Gd fuels” after their irradiation enables the performance of Gd as burnable poison to be evaluated. Major fission products of 235U that have a fission yield higher than 5 % are presented in Table 1.1 [26]. Although the percentage of fissile material in
SNF is very low, the nature of this material makes it a target for nuclear safeguards. Burn-up credit is an accurate and realistic means to determine SNF reactivity, because it takes into account the reduction in the reactivity of nuclear fuel during irradiation due to the net reduction of fissile nuclides and to the production of neutron-absorbing nuclides (fission products and non-fissile actinides) [27]. Activities related to the assessment of nuclear criticality safety in member countries of the NEA (Nuclear Energy Agency) are coordinated by the Expert Group on Burn-Up credit Criticality (EGBUC). SNF characterization (assays and burn-up credit) is carried out for criticality safety and safeguards purposes using non-destructive and destructive analysis (explained in sections 1.4.5 and 1.4.6 of this chapter) to determine the concentrations and isotopic compositions of the main fuel elements (uranium, plutonium and gadolinium when it is added as a burnable poison in “Gd fuels” [28]) and those of a fission product monitor (neodymium).
Figure 1.4 Composition of a UOx SNF with 3.5 % initial enrichment, 33 GWd.t(heavy metals)-1 and 3 years cooling time [1]
Table 1.1 Fission products of 235U with fission yield higher than 5 % [26]
Nuclide Fission yield (% per fission of 235U) ± U (k=2)
90Sr 5.73 ± 0.13
95Zr 6.502 ± 0.072
95Nb 6.498 ± 0.072
99Mo 6.132 ± 0.092
99Tc 6.132 ± 0.092
133I 6.59 ± 0.11
135I 6.39 ± 0.22
133Xe 6.6 ± 0.11
135Xe 6.61 ± 0.22
137Cs 6.221 ± 0.069
140Ba 6.314 ± 0.095
140La 6.315 ± 0.095
141Ce 5.86 ± 0.15
144Ce 5.474 ± 0.055
144Pr 5.474 ± 0.055
144Nd 5.475 ± 0.055
1.4.1 Uranium
Uranium has been formed within our Solar System by multiple supernovae from over 6 billion to about 200 million years ago [29]. As a result, uranium was present in the dust that eventually clogged together and cooled down to produce the Earth. Uranium is found naturally in trace amounts in different compartments of the Earth, such as soil (2 parts per million), water bodies (e.g. 0.3 parts per million in seawater [30]) and even air (2 µBq·m-3). According to the United Nations Scientific Council, worldwide average dose rates due to ingestion (mainly from water) and inhalation of natural uranium nuclides are less than 1 µSv per year [31].
Uranium was discovered in 1789 by Klaproth and its radioactivity was demonstrated by Becquerel in 1896. Natural uranium has 92 protons in its nucleus, its electronic configuration is [Rn] 5f3 6d1 7s2 and it has three naturally occurring isotopes: 234U, 235U and 238U, with characteristics as shown in Table 1.2. Nuclides with odd mass numbers (such as 235U in Table 1.2) generally have a larger neutron capture cross-section compared to nuclides with an even mass number, which makes the probability of neutron absorption higher for nuclides with odd mass numbers than for those with even mass numbers (Oddo-Harkins rule) [32]. Atomic nuclei with even mass numbers are more stable when formed than those with odd mass numbers. This phenomenon can be explained by the nuclear shell model [32].
Table 1.2 Characteristics of naturally occurring uranium nuclides (the characteristics listed in this table were sourced from Nucleonica [33], except for the relative abundances which were calculated based on mole fractions obtained from IUPAC [34])
Nuclide
Relative abundance
(wt. %)
Half-life (years)
Specific activity (Bq·g-1)
Neutron absorption cross-section
(b)
Decay mode
234U 0.00531 (49) 2.457·105 (3) 2.300·108 (3) 116 α
235U 0.71137 (59) 7.038·108 (5) 7.996·104 (6) 697 α
238U 99.2833 (16) 4.468·109 (3) 1.2436·104 (8) 12 α Due to its electronic configuration, uranium is most stable in oxidation state VI, and is mainly found as UO22+ (uranyl) in complexes [32]. The uranyl group can be detected in an infrared spectrum of a uranium compound by the presence of (1) a strong band in the region 920-980 cm-1 due to the asymmetric O–U–O stretching vibration or (2) a band around 860 cm-1 caused by the symmetric O–U–O stretching vibration in a Raman spectrum. Additionally, uranyl complexes have a yellow colour and a characteristic absorption peak around 25,000 cm-1 (400 nm) which can be monitored using UV-Vis spectrometry.
1.4.2 Plutonium
Plutonium is a man-made element first produced in 1940 by Seaborg et al. by bombarding uranium with neutrons [35]. Plutonium found in the environment originates from anthropogenic sources (nuclear weapon testing, nuclear accidents, etc.). Characteristics of plutonium nuclides are shown in Table 1.3. Plutonium can be present in SNF either as a product of neutron capture by 238U or it can be added to the fuel prior to its irradiation (MOx). Reactor-grade Pu is defined
as material composed of more than 18% of 240Pu [36]. The neutron cross-section of 240Pu (Table 1.3) limits its use in weapon grade Pu, which contains at least 93 % 239Pu [36].
Table 1.3 Characteristics of plutonium nuclides [33]
Nuclide Half-life (years) Specific activity (Bq·g-1)
Neutron absorption cross-section
(b)
Decay mode
238Pu 87.7 (3) 6.34·1011 (2) 584 α
239Pu 2.4114·104 (11) 2.2947·109 (10) 1,029 α
240Pu 6.563·103 (5) 8.396·109 (6) 290 α
241Pu 14.33 (4) 3.829·1012 (11) 1,383 β-
242Pu 3.735·105 (11) 1.463·108 (4) 30 α
244Pu 8.00·107 (9) 6.77·105 (8) 12 α
Due to their similar redox potentials: E(Pu(III)/Pu(IV)) = 0.98V, E(Pu(IV)/Pu(V)) = 1.04 V and E(Pu(V)/Pu(VI)) = 0.94 V [32], multiple oxidation states III, IV, V and VI of plutonium can be present simultaneously in acidic solution. In alkaline solutions, plutonium can be found in oxidation states VII and even VIII (> 1 M sodium hydroxide) [37]. The oxidation states IV and VI of plutonium are the most stable ones in acidic solutions. Due to its large charge-to-radius ratio, plutonium (VI) readily strips oxygen atoms from water molecules and is present as PuO22+
(plutonyl) which is stable in aqueous solutions [38].
1.4.3 Lanthanides
The lanthanide series consists of 15 elements in the periodic table from lanthanum to lutetium with atomic numbers from 57 to 71, and its elements are known to have very similar chemical properties, making their separation far from straightforward. Most lanthanides (including neodymium and gadolinium) are present in solution as Ln3+ions having [Xe] 4fn electronic configurations, nevertheless some lanthanides can also exist in solution as Ln4+ (such as Ce) and Ln2+ions (such as Eu) [32]. The lanthanides differ in the number of electrons in the 4f orbital. As this orbital is positioned near the atomic nucleus, the electrons in the 4f orbital have only a limited effect on chemical bonding characteristics and speciation. Separation strategies therefore make use of the phenomenon called ‘lanthanide contraction’: Across the lanthanide series from La to Lu, the atomic and ionic radii are known to decrease as a result of the increasing effective nuclear charge (as is always the case within a period of the periodic table), however, the 4f electrons are less efficient in shielding the electrons inside the 5s and 5p orbitals from the positive nuclear charge, which leads to the “lanthanide contraction” [32]. Hence, the lanthanides share such similar chemistries [including oxidation state] that the possibilities of using the selectivity of an ion exchanger as the basis of a separation only can be rather limited for lanthanides.
1.4.3.1 Neodymium
Neodymium was discovered in 1841 by Carl Gustav Mosander, who called it didymium (from
“didymos”, which means twin in Greek) due to its similarity to lanthanum, which he had
discovered two years previously. In 1885, Carl Aeuer von Welsbach separated didymium into two elements and called these neodymium and praseodymium, both names deriving from the Greek meaning new and green twin, respectively [34]. Neodymium is the second most abundant lanthanide in the Earth’s crust (40 parts per million) after cerium (66 parts per million) [32].
Each neodymium atom has 60 protons in its nucleus. Characteristics of naturally occurring Nd nuclides are listed in Table 1.4, which shows that neodymium nuclides are mostly stable (142,
143, 145 & 146Nd) or have very long half-lives (more than 1015 years).
Table 1.4 Characteristics of naturally occurring neodymium nuclides (the characteristics listed in this table were sourced from Nucleonica [33], except for the relative abundances which were calculated based on mole fractions obtained from IUPAC [34])
Nuclide
Relative abundance
(wt. %)
Half-life (years) Specific activity (Bq·g-1)
Neutron absorption
cross- section (b)
Decay mode
142Nd 26.712 (44) Stable - 27 -
143Nd 12.062 (27) Stable - 404 -
144Nd 23.743 (25) 2.29·1015 (16) 4.0·10-2 (3) 4 α
145Nd 8.332 (13) Stable - 60 -
146Nd 17.388 (35) Stable - 6 -
148Nd 5.903 (22) 2.7·1018 3.3120·10-5 (8) 12 2β-
150Nd 5.860 (29) 2.1·1019 (5) 4.2·10-6 (10) 6 2β- 1.4.3.2 Gadolinium
Gadolinium was discovered by Jean-Charles Galissard de Marignac in 1886 and was named by Paul-Emil Lecoq de Boisbaudran after gadolinite, the mineral in which it was found.
Characteristics of gadolinium nuclides can be found in Table 1.5. Gadolinium nuclides are stable, except for 152Gd and 160Gd, which have long half-lives (more than 1014 years). When used as a burnable poison in nuclear fuels, the fuel is enriched to 5-10 wt. % in gadolinia (Gd2O3). As such, Gd is present as a major element within such fuels. During irradiation, gadolinium nuclides with large neutron absorption cross-sections, such as 155Gd and 157Gd (Table 1.5), capture neutrons thereby forming 156Gd and 158Gd, respectively. These resulting nuclides, however, have small neutron absorption cross-sections and thus constitute a dead-end for this decay chain reaction. Additionally, gadolinium nuclides can be produced in nuclear reactors by the fission of 235U and by the decay of fission products, but with low total yields (<
10-4 %).
Table 1.5 Characteristics of naturally occurring gadolinium nuclides (the characteristics listed in this table were sourced from Nucleonica [33], except for the relative abundances which were calculated based on mole fractions obtained from IUPAC [34])
Nuclide
Relative abundance
(wt. %)
Half-life (years) Specific activity (Bq·g-1)
Neutron absorption
cross- section (b)
Decay mode
152Gd 0.193 (29) 1.08·1014 (8) 8.1·10-1 (6) 735 α
154Gd 2.134 (20) Stable - 85 -
155Gd 14.581 (91) Stable - 60,900 -
156Gd 20.297 (41) Stable - 1,5 -
157Gd 15.617 (45) Stable - 254,000 -
158Gd 24.946 (87) Stable - 2.2 -
160Gd 22.232 (43) 1.3·1017 6.3622·10-4 0.77 2β-
1.4.4 Other spent fuel components
While burn-up determination requires the measurement of specific nuclides to assess the performance of the nuclear fuel, SNF characterization involves the measurement of many more nuclides formed during the irradiation of the fuel and is used to evaluate safety codes. Table 1.6 lists commonly measured nuclides for different safety-related SNF applications including burn-up determination.
Table 1.6 Commonly measured nuclides for different safety-related SNF applications (based on [21])
Nuclide Half-life
(years) Burn-up Radiological safety
Waste management
SNF characterization
79Se 2.95·105 + +
95Mo Stable + +
90Sr 28.9 + + +
99Tc 2.111·105 + + +
101Ru Stable + +
106Ru 371.6 days + +
103Rh Stable + +
109Ag Stable + +
125Sb 2.7586 + +
129I 1.6·107 + +
133Cs Stable + +
134Cs 2.065 + +
135Cs 2.3·106 + +
137Cs 30 + + + +
139La Stable + +
142Nd Stable +
143Nd Stable + +
144Nd 2.29·1015 + +
145Nd Stable + +
146Nd Stable + +
148Nd 2.7·1018 + +
150Nd 2.1·1019 + +
Table 1.6 Commonly measured nuclides for different safety-related SNF applications (based on [21]) (continued)
Nuclide Half-life
(years) Burn-up Radiological safety
Waste management
SNF characterization
144Ce 284.9 days + + +
147Pm 2.623 + +
147Sm 1.06·1011 + +
149Sm Stable + +
150Sm Stable + +
151Sm 90 + +
152Sm Stable + +
151Eu Stable + +
153Eu Stable + +
154Eu 8.59 + + +
155Eu 4.753 + +
152Gd 1.08·1014 +
154Gd Stable +
155Gd Stable + +
156Gd Stable +
157Gd Stable +
158Gd Stable +
160Gd 1.3·1017 +
234U 2.457·105 + + +
235U 7.038·108 + + +
236U 2.342·107 + + +
238U 4.468·109 + + +
237Np 2.14·106 + + +
238Pu 87.7 + + + +
239Pu 2.4114·104 + + + +
240Pu 6.563·103 + + + +
241Pu 14.33 + + +
242Pu 3.735·105 + + +
241Am 433 + + + +
243Am 7370 + + +
242Cm 162.8 days + +
243Cm 29.1 + +
244Cm 18.1 + +
245Cm 8.5·103 + + +
1.4.5 Non-Destructive analysis
Non-destructive passive or active analysis methods can be used to determine SNF characteristics in a fast and accurate way using computational codes. Passive methods are based on spontaneous radiation emission from the fuel itself, whereas active methods rely on external radiation sources. Several non-destructive analysis devices based on the monitoring of gamma rays, neutron emissions or Cherenkov light exist. Non-destructive analysis methods can be used to determine the burn-up and other characteristics of SNF based on computer codes, such as ALEPH [39] which links the Monte Carlo code MCNPX [40] to the burn-up code ORIGEN 2.2 [41]. ALEPH was developed by Wim Haeck and Bernard Verboomen at SCK CEN. An
example of a passive method is the PYTHONTM device [42], developed as a collaboration between EDF (Electricité de France) and CEA (Commissariat à l’Energie Atomique et aux énergies alternatives), which is a combination of a total gamma measurement, a passive neutron measurement and an online depletion code resulting in an accurate burnup determination within
± 2 % at a 95 % confidence level (k=2) [42]. Other non-destructive analysis methods exist, such as self-indication neutron resonance densitometry (SINRD), which is based on neutron measurements having uncertainties as high as 25 %, and the partial defect tester (PDET), which is based on neutron and gamma-ray measurements to detect partial defects in the fuel assembly and can be used to derive a semi-quantitative estimation of the burn-up. A device combining active and passive methods is the NAJA device [43], which includes passive and active neutron measurements combined with an on-line depletion code and gamma-spectrometry. The NAJA device is able to automatically determine the nature of the fuel (fresh or irradiated, UOx or MOx), the presence and kind of neutron absorber, the initial enrichment in 235U for a fresh UOx assembly and make accurate burn-up determinations with expanded uncertainties of ± 2 % at a 95 % confidence level (k=2) [43]. Neutron resonance densitometry (NRD) is an experimental active non-destructive method relying on neutron time-of-flight techniques, aiming to quantify uranium and plutonium nuclides in SNF in less than 20 minutes with less than 1 % resulting uncertainty [44].
The accuracy of the results obtained using computer codes is important in establishing the safety basis in SNF management [45]. Despite being a fast way to determine SNF characteristics, the results obtained by computational codes in non-destructive analysis must be validated by destructively analysing SNF samples using radiochemical analysis methods (explained in section 1.4.6 of this chapter). Thus, the accuracy of the results obtained using computer codes is tested by comparing experimental results obtained from destructive chemical analyses on real SNF samples with the corresponding results of the computer codes.
1.4.6 Destructive analysis
Interest in code validation using destructive analysis methods was acted upon by EGBUC in 2006, after a workshop on “The need for post-irradiation experiments to validate fuel depletion calculation methodologies” (held in Rez, Czech republic on May 11th-12th, 2006), with the establishment of a new expert group on assay data for spent nuclear fuel (EGADSNF). This expert group coordinates assay data activities and facilitates cooperation between NEA member countries in developing and implementing burn-up credit methods, taking into consideration the high cost of initiating new experimental assay programs (fuel transportation, hot-cell facilities, radiochemical analysis capabilities and requirements for waste management) [45].
The results obtained using destructive and non-destructive analysis methods are compared to validate computational codes. In Europe, SNF assay measurement and burn-up determination are conducted in several radiochemical analysis laboratories, including the Belgian Nuclear Research Centre (SCK CEN) in Belgium, the CEA in France, the European Commission’s Joint Research Centre (JRC) in Germany and the Paul Scherrer Institute (PSI) in Switzerland.
Destructive analysis remains the most reliable analytical approach to determine nuclide-specific concentrations in SNF and can only be performed at specialised laboratories with hot-cells and radiometric or mass spectrometric analysis equipment [45].
SNF assay can be carried out using hybrid K-edge densitometry (HKED) to determine the elemental concentrations of uranium and plutonium from 0.5 mg·L-1 to several hundreds of g·L-1 with uncertainties ranging from 2 to 10 % (k = 2) respectively. Figure 1.5 shows the HKED device used at CEA [46]. The HKED device uses results from two measurements performed simultaneously on the sample: K-edge transmission and X-Ray fluorescence [46].
Figure 1.5 HKED device used at CEA [46]
The uranium elemental concentration in SNF can also be accurately determined using the titration method developed by Davies and Gray in 1964. No separation is required prior to the assay of uranium in SNF using the Davies and Gray titration unless interferences are present in milligram amounts or larger. First, uranyl in the sample is reduced to oxidation state IV by using excess Fe(II) in concentrated phosphoric-sulfamic acid, whilst the excess Fe(II) is oxidized selectively by nitric acid in the presence of Mo(VI) as a catalyst. Then, using potentiometric titration with vanadium as an electrochemical enhancer, U(IV) is titrated using a K2Cr2O7 solution. The mass fraction of uranium (CU) in the sample can then be determined by using eq. 1.3 [47], where W is the atomic mass of uranium, T is the titer of the titrating solution used (in equivalent·g-1), mc is the mass of the titrating solution used (in grams) to reach the endpoint and ma is the mass of the sample used (in grams).
𝐶𝑈 = 𝑊𝑇 (𝑚𝑐
2𝑚𝑎) (eq. 1.3)
The Davies and Gray titration is a fast, accurate and precise method to determine the uranium concentration with relatively little sample manipulation required. For SNF characterization, the titrator has to be installed inside a hot-cell.
The SNF burn-up is determined by using destructive analysis to measure the nuclide-specific concentration of a fission product monitor, usually 148Nd [25], and those of the residual heavy
metal elements of the fuel: U and Pu [47-49]. Several destructive analysis methods are available, such as alpha spectrometry, liquid scintillation counting, gamma spectrometry, thermal ionization mass spectrometry (TIMS) and inductively coupled plasma-mass spectrometry (ICP-MS) [45]. SNF burn-up determination requires combining some of these radiometric and mass spectrometric methods.
For burn-up determination using destructive analysis, SNF sample preparation takes place inside a hot-cell where SNF pellets are dissolved, most commonly in 8 to 10 M nitric acid heated near its boiling point (86 °C) under reflux, to transfer uranium, the fission products, and most of the plutonium and the minor actinides into the solution. The nitric acid solution is then filtered and the residue is treated using a mixture of 8 M nitric acid and 0.1 M hydrofluoric acid to dissolve all the plutonium oxide. The filtered solution is then combined with the second solution before further gravimetric dilution in 1 M nitric acid to reduce the dose rate to a level permitted in laboratories, before removing it from the hot-cell. After dissolving the SNF sample and diluting the corresponding digest, chromatographic separation (off-line or on-line) is used to isolate the main fission products (neodymium, samarium, europium, gadolinium), uranium, plutonium and minor actinides into pure fractions prior to determining the nuclide-specific compositions and elemental concentrations of these elements by radiometric or mass spectrometric methods. Amongst the destructive analysis mass spectrometric methods, coupling on-line chromatography is only possible with ICP-MS. Such a hyphenated approach provides a rapid separation of the elements present in the SNF from one another, has a higher sample throughput, generates less radioactive waste and exposes the operator to less radiation compared to off-line chromatography.
1.4.6.1. Radiometric methods
SNF contains various alpha-, beta- and gamma-emitting radionuclides that can be determined by using radiometric methods such as alpha spectrometry, gamma spectrometry and liquid scintillation counting. These radiometric methods require chromatographic separation of SNF elements into individual fractions to eliminate interfering radiation energies and to improve the limit of detection [45].
For measurement of short-lived alpha-emitting radionuclides in SNF, alpha spectrometry most commonly uses a passivated implanted planar silicon (PIPS) detector, which is placed inside a vacuum chamber, to absorb the high energy of an alpha particle (2-8 MeV) and convert it into an electronic signal (counts). PIPS detectors are usually calibrated for a range of 0 to 10 MeV by using standard alpha-emitting sources with known activities. Alpha spectrometry can be used to detect and quantify short-lived alpha-emitting radionuclides of uranium, plutonium, americium and curium in SNF with expanded relative uncertainties as low as 2 % in the best case at a 95 % confidence level (k = 2) [45]. However, alpha-spectrometry is incapable of separating the overlapping alpha-peaks of 239Pu (5105, 5144 and 5157 keV) and 240Pu (5124 and 5168 keV). Nevertheless, alpha spectrometry remains the method of choice to quantify shorter-lived radionuclides (e.g. 238Pu) when the plutonium is not completely separated from the highly concentrated uranium in SNF, which causes isobaric overlaps and hinders mass spectrometric analysis. Additionally, mass spectrometric methods, such as ICP-MS, can
